Page 453 - Dust Explosions in the Process Industries

420 Dust Explosions in the Process Industries

by which continuous, dense showers of sparks are produced and comparativelylarge hot
spots may be generated. The second is single accidental impacts.

5.4.2
SPARKS AND HOT-SPOTS FROM RUBBING,
GRINDING, AND MULTIPLE IMPACTS

The ability of metal sparks or hot spots from grinding operations to ignite dust clouds
has been demonstrated by several researchers. The experiments by Leuschke and Zehr
(1962) are probably the first ones in which dust clouds were ignited by grinding wheel
metal sparks. However, no clouds of organic dusts ignited. Zuzuki, Takaoka, and Fujii
(1965) ignited different coal dusts using both sparks and hot spots from a piece of steel
in contact with a grinding wheel rotating at 23-47 m/s peripheral velocity. Allen and
Calcote (1981) conducted similarexperiments,in which metal sparks were generated by
pressing a steel rod against a rotating grinding wheel. By retarding and focusingthe spark
stream, it was possible to ignite clouds of natural organic dusts such as corn starch and
wheat grain dusts.
Kachan et al. (1976) studied the ignition of clouds of coal dust by metal sparks or hot
spots generated by the cutters of a coal cutting machine, when cutting pyrite and sand-
stone at a speed of 1.5-2.0 ds. The coal dust contained 24% volatiles or more, and 85%
was finer than 75 pm. The dust concentrationwas 100g/m3.Using pyrite containing more
than 35% sulfur, and with a load per cutter of 1-3 kW, the probability of ignition was
practically 100%.However, the coal dust cloud did not ignite until after 15-80 s of con-
tinuous cutting with sparking, depending on the load. This long delay suggests that the
ignition source was not the spark shower but a hot spot generated either at the cutter tip
or on the pyrite surfacejust behind the cutter.
Ritter (1984a, 1984b) and Muller (1989) conducted extensive studies of ignition of
dust clouds by sparks and hot surfaces generated by scratching, grinding, and multiple-
impact processes. They used the concept of equivalent electric spark energy for char-
acterizing the ignition potential of the various scratching, grinding, and impact sources
studied. This was done by first determining the lowest concentration of a given dust in
air at which an essentially quiescent dust cloud could be ignited by the heat sourceinves-
tigated. The minimum electric spark ignition energy at this particular dust concentra-
tion was then determined and taken as the equivalent electric spark energy of that
particular heat source.
Ritter and Muller found linear correlations between the minimum ignition tempera-
ture of the dust cloud determined by the BAM furnace (Chapter 7) and the logarithm of
the equivalent minimum electric spark ignition energy, for the various ignition sources
investigated, as illustrated in Figure 5.23.
The example indicated by dot-dashed lines says that a dust cloud of minimum igni-
tion temperature 615°C cannot be ignited by flint sparks from grinding or scratching
unless its minimum ignition energy is lower than about 20 mJ. Similarcorrelationswere
found for sparks and hot spots from multiple impacts.
Unfortunately the type of relationships illustrated in Figure 5.23 are not generally
applicable because in practice grinding, scratching, and impact processes may differ